[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]

analytical uncertaity of pCO2 and carbon chem.




Dear Colleagues,

After following the discussion about issues of carbon chemistry and chemical
equilibrium constants by Rick, Ernst, and others in the context of
OCMIP and GOSAC it appears that you might be
interested in a MS that we (R. Fink, R. Murnane, F. Joos) have submitted to
TELLUS. I was very happy to see Rik's data that will help to resolve
controversial issues regarding equilibrium constants and pCO2 data quality.

Our MS is about analytical uncertainties of surface water pCO2
measurements. Two different temperature correction
methods are compared. The temperature correction accounts for differences between
the in-situ temperature of the sample and the temperature of the
equilibrator at which mole fraction measurements of CO2 are
performed. The first correction
method uses a chemistry model based on the constants as published by
Roy et al., Goyet and Poisson, and Millero. The second method uses an
exponential relationship as published by Takahashi. We find that for
large temperature differences the two methods yield systematic
deviations up to 30 ppm. Obviously, both methods cannot be true.

We suggest that the discrepancy between the two
correction methods should be resolved in order to increase our
confidence in available equilibrium constants and may be more
important the confidence in available surface water pCO2 measurements
that have undergone such a temperature correction. 

With respect to modeling, we do not expect that eventual revision of
equilibrium constants and their temperature dependency would seriously
affect prediction of anthropogenic CO2 uptake obtained by ocean
transport models.  We find that the increase in pCO2 divided by the
increase in DIC (Delta-pCO2 / Delta-DIC) of surface water is within a
range of 13% at 0 Celsius, 8% at 10 Ccelsius, and of 5% at 20 Celsius
and 30 Celsius when emplyoing 6 different sets of equilibrium
constants
\shortcite{Buch_51,Edmond_gca_70,Goyet_dsr_89,Hansson_dsr_73,Lyman_1956,Mehrbach_lo_73}
at a CO2 partial pressure of 280 and 400 ppm. Differences between the
constants of Goyet and Poisson and those of Mehrbach are
smaller. In any case, I support the use of one set of constants when
comparing different models.

You will find the abstract of the submitted version attached. The whole
manuscript can be downloaded from my www page as a postscript file:

http://www.climate.unibe.ch/~joos/OUTGOING/publications/fink_murnane_joos_tel_97

If you would prefer to have a hard copy version sent to you, or in case
of difficulties, please contact me.

With best regards,

Fortunat



ABSTRACT:

The analytical uncertainties in the CO2 fugacity fCO2 of surface sea
water are investigated using a chemistry model of the carbonate system
and Gaussian error propagation. We find a typical 2-sigma uncertainty
in fCO2 of 1% if the fugacity is determined by mole fraction
measurements on air equilibrated with the sample at or close to
in-situ temperature. Frequently, equilibrated air is analyzed at a
constant temperature of 25, 20, or 4 Celsius and it proves necessary
to correct for differences between in-situ temperature and the
temperature of the equilibrated air. A comparison between our full
chemistry calculation and an exponential temperature correction method
\shortcite{Takahashi_gbc_93} shows systematic deviations as large as
30 uatm when correcting for temperature differences up to 20 Celsius
as exploited for Antarctic samples \shortcite{Chipman_cdiac_94}. The
exponential correction always yields larger fCO2 at in-situ
temperature if the mole fraction measurements were done at 20
Celsius. Deviations are largest for high-latitude samples. Our results
imply (1) that published estimates of the air-sea partial pressure
difference could be seriously biased depending on the temperature
correction method applied; (2) that published estimates of global
air-sea net carbon fluxes \shortcite{Tans_sci_90,Takahashi_pnas_97}
may be too low by as much as a GtC/yr ; (3) or ,alternatively, that
published chemical equilibrium constants may not be adequate for
correction.
- -- 
- ----------------------------------------------------------------------------
Fortunat Joos
Physics Institute, KUP, Sidlerstr. 5, CH-3012 Bern

Phone:    ++41(0)31 631 44 61
Fax:      ++41(0)31 631 44 05
e-mail:   joos@climate.unibe.ch
Internet: http://www.climate.unibe.ch/~joos/